Development of a New Posture Evaluation Method Using Image Recognition Technology by Machine Learning

2023 Fiscal Year Final Research Report

• Project/Area Number: 21K11510
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Review Section: Basic Section 59030:Physical education, and physical and health education-related
• Research Institution: Kyoto Tachibana University
• Principal Investigator: Kai Yoshihiro 京都橘大学, 健康科学部, 准教授 (90632852)
• Co-Investigator(Kenkyū-buntansha): 村田 伸 京都橘大学, 健康科学部, 教授 (00389503) ; 来田 宣幸 京都工芸繊維大学, 基盤科学系, 教授 (50452371)
• Project Period (FY): 2021-04-01 – 2024-03-31
• Keywords: 姿勢評価 / 画像認識 / 運動器疾患 / 身体機能 / 加齢変化

Outline of Final Research Achievements

Changes in posture due to aging are one of the indicators of the health status of elderly individuals. In this study, we have developed an evaluation system using image recognition technology based on machine learning, enabling non-medical personnel to easily detect posture changes. The subjects were community-dwelling elderly individuals, and their static standing postures were photographed from the sagittal plane using a digital camera. Additionally, the subjects' postures were classified into several categories by physical therapists. Gray-scale and silhouette images were generated from the photographed posture images, and the accuracy of identifying these images was examined. The analysis results showed that both the gray-scale and silhouette images could generally provide judgments similar to those made by the physical therapists.

Free Research Field

応用健康科学

Academic Significance and Societal Importance of the Research Achievements

従来より、運動器疾患を専門とする整形外科やリハビリテーション分野では、筋骨格の老化度の予測や関節に加わる機械的ストレスの推定、運動障害のメカニズムを予測する指標として、姿勢変化の評価が幅広く用いられてきた。しかしながら、姿勢の変化を客観的に捉えるためには、高価な測定装置や熟練の専門家による評価が必要であり、専門家以外が簡便にかつ客観的に姿勢を評価する術はなかった。専門家でなくとも、姿勢や動作の変化を簡便かつ客観的に評価でき、その結果を筋骨格の老化の指標として活用することができれば、一次予防の観点からも、運動習慣の定着やその行動変容につながる実効的な対処を先制して実施できる可能性がある。